This research was conducted to verify quantitatively the effects of "Data-driven production", which depends on both the integrated agricultural datasets and their analysis to obtain higher yield of next year, in large-scale rice production. Datasets contains yield of each field measured by combine, and data collected from all cultivation practices. Data were collected at farm corporation, located in southern part of Ibaraki Prefecture, approximately 160 ha in total, growing its scale through "extension of both transplanting and harvesting duration through the combination of multiple cultivars", during 5 years since 2019, when the farm participated the research project of "smart agriculture acceleration demonstration" funded by Ministry of Agriculture, Forestry and Fisheries through 2 years, until 2023 with 3 years of follow-up. Major improvements and their results were as follows: Earlier transplanting orders were allocated to a low-amylose cultivar and a glutinous cultivar, those are sold at higher price, resulting in stable yield due to better climate condition during ripening. Fields of 'Koshihikari' were aggregated to the block where large-scale fields exist, resulting in faster machine operation. Pesticide for stink bugs was applied to the early-maturing cultivars, whose panicles emerged earlier than nearby fields, since 2020. Lodging-resistant late-maturing cultivars were chosen to the late-transplanting block due to the water constraints, and high fertilization of twice topdressing by unmanned aerial vehicle (UAV) stabilized their yields. The increase of yield, resulting in over 110% level of common annual yield in southern Ibaraki Prefecture, was proven through these improvements of cultivation practices, based on analysis of comprehensively collected data.
View full abstract